2Department of Biochemistry
3Department of Physiology, Development and Neuroscience University of Cambridge, Cambridge CB2 1QR, UK
4Human Biology Division, Fred Hutchinson Cancer Research Centre, Seattle, WA 98109, USA *Correspondence:[email protected](J.B.),[email protected](A.S.)
http://dx.doi.org/10.1016/j.cell.2013.03.012
SUMMARY
Factors that sustain self-renewal of mouse
embry-onic stem cells (ESCs) are well described. In contrast,
the machinery regulating exit from pluripotency is ill
defined. In a large-scale small interfering RNA
(siRNA) screen, we found that knockdown of the
tumor suppressors
Folliculin (Flcn) and Tsc2 prevent
ESC commitment. Tsc2 lies upstream of mammalian
target of rapamycin (mTOR), whereas Flcn acts
downstream and in parallel. Flcn with its interaction
partners Fnip1 and Fnip2 drives differentiation by
re-stricting nuclear localization and activity of the bHLH
transcription factor Tfe3. Conversely, enforced
nuclear Tfe3 enables ESCs to withstand
differentia-tion condidifferentia-tions. Genome-wide locadifferentia-tion and
func-tional analyses showed that Tfe3 directly integrates
into the pluripotency circuitry through transcriptional
regulation of
Esrrb. These findings identify a
cell-intrinsic rheostat for destabilizing ground-state
plu-ripotency to allow lineage commitment. Congruently,
stage-specific subcellular relocalization of Tfe3
sug-gests that Flcn-Fnip1/2 contributes to
develop-mental progression of the pluripotent epiblast in vivo.
INTRODUCTIONStem cell fate is determined by a balance between pro-self-renewal and pro-differentiation signals. To initiate differentiation, stem cells have to be forced out of self-renewal, the trans-criptional networks conferring stem cell identity need to be dissolved, and lineages have to be chosen. How these interde-pendent tasks are coordinated is poorly understood. Mouse embryonic stem cells (ESCs) provide a tractable system for dis-secting this process because intrinsic and extrinsic signals regu-lating self-renewal are reasonably well defined.
Genetic and biochemical studies have shown that activity of mitogen-activated protein kinase (MAPK) (Kunath et al., 2007)
and glycogen synthetase kinase-3 (GSK3) (Doble et al., 2007) prime ESCs for differentiation. Chemical inhibition of both (2i) promotes robust self-renewal (Ying et al., 2008). It is therefore hypothesized that shielding ESCs from differentiation cues is sufficient for stabilization and propagation of a naive pluripotent ground state (Nichols and Smith, 2009).
Pluripotency is governed by a transcription factor network that contains numerous autoregulatory loops, suggestive of self-propagation (Jaenisch and Young, 2008). How this recur-sive circuitry is extinguished in an orderly manner to enable developmental progression and eventual lineage commitment is currently unclear. Endogenous repressors Tcf3 (encoded by Tcf7l1) and the NuRD complex (Kaji et al., 2006; Nichols and Smith, 2012; Pereira et al., 2006) play a role, but other key drivers of exit from naive pluripotency have yet to be identified.
Delineation of mechanisms that steer ESCs into differentiation may reveal predetermined breaking points in the pluripotent transcription factor network. Furthermore, factors regulating ESC fate transition are candidates to drive progression of the naive preimplantation into the lineage-poised postimplantation epiblast in vivo (Nichols and Smith, 2009).
Here, we exploited the ground-state culture system to screen functionally for genes promoting exit from naive pluripotency.
RESULTS
A Large-Scale siRNA Screen Identifies Genes Required for Exit from Ground-State Pluripotency
To uncover genes that are required for progression from naive pluripotency into a state primed for differentiation, we used a simple monolayer differentiation protocol (Ying et al., 2003). Effi-cient loss of ESC identity over 3 days (Figures 1A and 1B) is a time window compatible with perdurance of transient knock-down by siRNA. We assayed for resistance to exit the ESC state by ability to proliferate in selective 2i culture conditions plus re-tained expression of the Oct4 locus. After transfecting siRNAs in 2i, differentiation was enabled by withdrawal of the inhibitors. Resistance to commitment was then assayed by reapplying inhibitors and selecting for Oct4 expression (Figure 1A). After Open access under CC BY license.
72 hr, cells transfected with GFP, Klf4, or Nr0b1 siRNAs lost ESC properties, but upon knockdown of FGF4, GSK3a/b, or Tcf3, sig-nificant numbers of alkaline phosphatase (AP) expressing ESCs were recovered (Figure 1B).
To allow quantitation at high throughput, commitment resis-tance was measured by cell survival. We screened roughly 9,900 genes in duplicate experiments with pools of four indepen-dent siRNAs. Z scores were determined for each run (Figure 1C andFigure S1A available online). Seventy genes that stringently scored positive (Z > 3 and Z > 2.5 in the two trials) were inter-sected with RNA-sequencing data from ESC differentiations (T. Kalkan and A.S., unpublished data), and 17 genes were excluded due to lack of expression. Another 7 genes were manually eliminated as likely false-positives (see Extended Experimental Procedures). The remaining 46 primary hits were retested, and 28 validated with at least two different siRNAs ( Fig-ure S1B). These include Tcf3 and other members of the path-ways inhibited in 2i culture conditions (Figure 1D), indicating that the screen successfully identified genes regulating exit from ESC pluripotency.
Folliculin Regulates Exit from Pluripotency
We focused on the tumor suppressor Folliculin (Flcn) (Nickerson et al., 2002) because a role in pluripotency has not been described, but homozygous knockout alleles show early embry-onic lethality (Chen et al., 2008a; Hasumi et al., 2009). Rex1GFPd2 ESCs that quantitatively report differentiation (Marks et al., 2012;Wray et al., 2011;Yang et al., 2012) were engineered to express a small hairpin RNA (shRNA) targeting Flcn, resulting in reduced transcript and protein (Figures S2A and S2B). Twenty-four hours after 2i withdrawal, control cells downregulated GFP, whereas Flcn shRNA knockdown lines maintained expression (Figure 2A). This phenotype could be partially reverted by a Flcn transgene designed to be shRNA resistant (Figure S2C). Maintenance of GFP expression at 24 hr without 2i was also observed after transient knockdown with siRNA transfection (Figures S2D and S2G). Replating differenti-ating control cells after 3 days into 2i conditions with Rex1-expression selection gave very few AP-positive colonies, whereas Flcn knockdown cells yielded many (Figure 2B). We used colony-forming assays to quantify 10%–15% of cells in
Figure 1. A Large-Scale siRNA Screen for Genes Regulating ESC Commitment
(A) Outline of the screening procedure; O4GIP ESCs were transfected with siRNAs in 2i, differentiation enabled by inhibitor removal, and resistance to commitment assayed by restoring 2i with puromycin selection for Oct4 expression.
(B) Exit from pluripotency in differentiating O4GIP ESCs transfected with indicated siRNAs assayed after 24 hr, 48 hr, and 72 hr and stained for AP. (C) Average screen Z scores. Red and green triangles show validated hits (seeFigure S1B); gray triangles show duplicates within the transcription factor subset. (D) O4GIP ESC resistance to commitment after transfection with siRNAs was quantified with a cell viability assay and normalized to no siRNA transfection controls. Pools and individual siRNAs are shown. Note that Tsc1 was not recovered in the primary screen. For siRNA pools, the average and standard deviation (SD) of two technical replicates is shown.
Flcn shRNA knockdown cell lines as uncommitted 3 days after inhibitor withdrawal (Figure 2C). This is a conservative estimate because cloning efficiency of ESCs is typically only 50% (Wray et al., 2010). The phenotype could be rescued with the transgene (Figures 2D and S2E). Resistance to differentiation was also observed in the presence of Activin A, fibroblast growth factor 4 (FGF4), and bone morphogenetic protein 4 (BMP4) or 10% fetal bovine serum (Figure 2B), suggesting that Flcn is important for exit from pluripotency under different inductive regimes.
To confirm these findings in a genetic null background, we derived ESCs from Flcnflox/+intercrosses (Hasumi et al., 2009). Cells were stably transfected with a plasmid expressing Cre fused to the ligand-binding domain of the estrogen receptor (ERT2), and single-cell clones expanded. After treatment with 4-Hydroxytamoxifen (Tam) to activate Cre, Flcn protein was undetectable in two independently derived knockout ESC lines (Figure S2F). After 3 days of differentiation, wild-type and hetero-zygous cells lost colony-forming ability, whereas 20%–30% of the knockout cells generated undifferentiated colonies ( Fig-ure 2E). These observations confirm that Flcn expedites exit from the ESC state.
Folliculin Acts downstream of, or in Parallel to, mTOR and Together with Fnip1/2
Flcn has been linked to mammalian target of rapamycin (mTOR) kinase signaling (Baba et al., 2006;Hartman et al., 2009;Piao et al., 2009). The mTOR inhibitor Tsc2 scored as a commitment regulator in our screen (Figure 1D). Tsc2 functions as a hetero-dimer with Tsc1 (Laplante and Sabatini, 2012), and knockdown of Tsc1 also inhibited exit from pluripotency (Figure 1D). How-ever, we found that mTOR was strongly activated after inhibitor withdrawal, leading to increased phosphorylation of ribosomal protein S6 (S6), S6kinase, and the translational regulator 4EBP1 (Figure 3A). The phosphorylation kinetics differed; in particular, S6 and 4EBP1 phosphorylation peaked at 24 hr and 48 hr, respectively. These effects are not due to starvation in N2B27 as they were also observed in serum. To test whether activation of mTOR is important for ground-state exit, we quan-tified undifferentiated ESCs at 24 hr, 36 hr, and 48 hr of 2i withdrawal in the presence or absence of the mTOR inhibitor rapamycin (Rapa) (Figure 3B). No difference was detected. As expected, knockdown of Tsc2 increased S6 phosphoryla-tion (Figure S3A) and impaired Rex1GFPd2 downregulation
C D E
Figure 2. Flcn Regulates ESC Commitment
(A) Flow cytometry profile of Rex1GFPd2 expression in Flcn shRNA knockdown clones (Flcn.2,4) and controls (ctrl.1,2) in 2i conditions (left panel) and after 24 hr of 2i withdrawal (right panel).
(B) Flcn shRNA cell lines and controls kept in 2i (10-fold fewer cells) or differentiated for 72 hr in N2B27, or N2B27 supplemented with 10% FBS or 25 ng/ml FGF4, 10 ng/ml BMP4, and 20 ng/ml Activin A, were replated in 2i and selected in blasticidin for Rex1 expression. Resulting ESC colonies were visualized by AP staining. (C and D) Differentiating Flcn shRNA cell lines (C) or Flcn shRNA cell lines expressing an empty vector or shRNA-resistant Flcn transgene (rescue) (D) and respective controls were replated at clonal density, and colonies arising from uncommitted cells stained for AP. Average clonogenicity and SD are relative to number of plated cells of two independent experiments.
(E) CreERT2-expressing clones of indicated genotypes were treated with Tam, differentiated for 3 days, and uncommitted cells quantified in 2i/LIF. Average and SD are of at least three independent biological replicates.
(Figure S2D). Blocking mTOR with Rapa in this context abro-gated the differentiation defects (Figure 3C). These data indicate that elevated activity of mTOR in ESCs impedes exit from self-renewal, but that activation during differentiation is not required for cell-fate transition. Rapa did not affect Tcf3 or Flcn knock-down phenotypes (Figure 3C), suggesting that Flcn acts down-stream or independent of mTOR.
Flcn has been reported to bind to two Flcn-interacting pro-teins, Fnip1 and Fnip2 (Baba et al., 2006;Hasumi et al., 2008;
Takagi et al., 2008). To identify binding partners in ESCs, we immunopurified (IP) Flcn from cells expressing functional 33FLAG-Flcn (data not shown). Silver staining revealed the presence of one specific copurifying protein at approximately 130 kDa that was identified as Fnip1 by mass-spectrometry ( Fig-ure S3B, arrowhead). Vice versa, 33FLAG-Fnip1 copurified one specific protein with a molecular weight similar to that of Flcn (Figure S3B, open arrowhead). Endogenous Fnip1 and Flcn were detected in 33FLAG-Flcn and 33FLAG-Fnip1 IPs, respec-tively (Figure 3D), confirming the presence of a Flcn-Fnip1 pro-tein complex in ESCs.
Flcn, Tsc1, Tsc2, and Fnip1 messenger RNA (mRNA) levels did not change markedly during early differentiation as compared to downregulation of Nanog and upregulation of FGF5 (Figure 3E). Furthermore, Flcn or Fnip1 overexpression did not induce transcriptional changes indicative of differentia-tion (Figure S3C) or accelerate ESC commitment (Figure S3D).
This suggests that Flcn and Fnip1 levels are not limiting for exit from pluripotency.
To test the functional significance of the Flcn-Fnip1 interac-tion, we knocked down Fnip1 and its homolog Fnip2 and assayed loss of ESC identity after 2i withdrawal. Although neither Fnip1 nor Fnip2 single knockdown had a strong impact, knock-down of both had a comparable effect to loss of Flcn (Figures 3F and S3E). Consistent with this, Rex1GFPd2 cells transfected with Fnip1/2 siRNAs retained GFP expression after 24 hr of with-drawal from 2i (Figure S2D).
To investigate whether the functions of mTOR and Flcn in regulation of pluripotency are ESC specific, we turned to postim-plantation epiblast stem cells (EpiSCs) (Brons et al., 2007;Tesar et al., 2007). EpiSCs express Oct4, Flcn, Fnip1/2, Tcf3, and Tsc1/ 2 (Hayashi et al., 2011). O4GIP EpiSCs were induced to differen-tiate for 3 days through withdrawal of growth factors and assayed for commitment by reapplying EpiSC culture conditions and selecting for Oct4 expression. Neither Tcf3, Tsc2, Flcn, nor Fnip1/2 knockdown impaired EpiSC differentiation (Figure S3F), suggesting that the role of these factors is specific to the cell context.
Flcn RegulatesTfe3 Localization and Activity
Flcn regulates nuclear translocation of the bHLH transcrip-tion factor Tfe3 in human kidney cells and mouse embryonic fibroblasts (Hong et al., 2010). Tfe3 showed heterogeneous
A D
B
C F
E
Figure 3. Flcn Acts downstream of or in Par-allel to mTOR and Interacts with Fnip1/2 (A) Rex1GFPd2 cells were differentiated in N2B27 with and without 20 nM Rapa (left panel) or in N2B27/10%FBS (right panel), and cell lysates probed with indicated antibodies.
(B) Rex1GFPd2 cells differentiated in N2B27 with and without 20 nM Rapa were replated at single-cell density in 2i including Rex1-expression selection at the indicated time points. The average percentage of uncommitted cells forming AP-positive colonies relative to the number of cells plated and SD are of two technical replicates. (C) Rex1GFPd2 cells transfected with indicated siRNAs were differentiated for 72 hr in N2B27 with and without 20 nM Rapa and replated in 2i with Rex1-expression selection, and resulting colonies were stained for AP.
(D) Proteins were immunoprecipitated with FLAG antibodies from stably transfected Rex1GFPd2 cells cultured in 2i or differentiated for 40 hr and probed with indicated antibodies.
(E) mRNA levels were quantified during differenti-ation and normalized to 2i-cultured cells. Average and SD are of two cell lines.
(F) O4GIP ESCs were transfected with indicated siRNAs, and after differentiation for 3 days, exit from pluripotency quantified with a cell-viability assay and normalized to no siRNA transfection controls. Average and SD are of two technical replicates.
subcellular localization in ESCs in 2i and could be detected in the nucleus (Figure 4A, left panel, arrowhead), cytoplasm (Figure 4A, left panel, open arrowhead), or both compartments. Staining
was specific as it was reduced in Tfe3 shRNA knockdown cells (Figure 4A, lower panel). Tfe3 mRNA levels did not change distinctly during ESC differentiation (Figure 3E), but we observed
C D
E
F
Figure 4. Flcn Regulates Subcellular Localization of Tfe3
(A) Control (ctrl.1), Flcn shRNA (Flcn.4), and Tfe3 shRNA (Tfe3.3) cell lines were stained for Tfe3 and DNA in 2i conditions (left panel) and 24 hr after inhibitor withdrawal (right panel). Tfe3 was detected in the nucleus (arrowhead) and cytoplasm (open arrowhead).
(B) Cytoplasmic (C) and nuclear (N) fractions of control and Flcn shRNA cells probed with indicated antibodies (open arrowheads indicate Fnip1 and Flcn bands; both bands recognized by Tfe3 antibodies are specific and likely represent phosphorylation variants;Hong et al., 2010).
(C) Box and whisker plots of nuclear/cytoplasmic Tfe3 ratios in ESCs transfected with indicated siRNAs in 2i conditions and 24 hr after inhibitor withdrawal. Indicated cell numbers (white) from three experiments with two ESC lines were quantified. (**) and (*) indicate Student’s t test values < 13 10100
and 13 1050 , respectively.
(D–F) Tfe3 is localized to the nucleus (arrowhead) at E3.5 (D). At E4.5, Tfe3 is found in the nucleus and cytoplasm of Nanog-positive epiblast cells (arrowhead) but stays nuclear in GATA4-positive (open arrowhead) primitive endodermal cells (E). At E5.5, Tfe3 is enriched in the cytoplasm (arrowhead) of Oct4-positive epiblast cells and remains nuclear (open arrowhead) in extraembryonic endoderm cells (F).
a marked cytoplasmic relocalization (Figure 4A, right panel). In contrast, most Flcn shRNA cells featured exclusively nuclear Tfe3 in 2i (Figure 4A, left panel), and many retained nuclear pro-tein after inhibitor withdrawal (Figure 4A, right panel, arrowhead). Consistently, biochemical nucleocytoplasmic fractionations showed more Tfe3 in the nuclear fraction of Flcn shRNA knock-down cells compared to control ESCs (Figure 4B). Tfe3 was also enriched in the nuclei of Flcn null ESCs (Figure S4A). We used automated image quantification to determine nuclear-cyto-plasmic ratios of Tfe3 in siRNA-treated cells (see Extended Experimental Procedures). Compared to controls, knockdown of Flcn and Fnip1/2 lead to a significant increase in nuclear Tfe3 both in 2i conditions and 24 hr after inhibitor withdrawal (Figure 4C). Importantly, this was not an indirect consequence of maintained ESC identity, as Tcf3-depleted cells showed cyto-plasmic Tfe3. Knockdown of Tsc2 increased nuclear Tfe3 con-centration in an mTOR-dependent manner (Figures 4C and
S4B), although less pronounced than after knockdown of Flcn or Fnip1/2. These observations suggest that Flcn-Fnip1/2 and Tsc2 activities converge on subcellular localization of Tfe3 and that nuclear exclusion of Tfe3 normally correlates with loss of ground-state pluripotency.
In EpiSCs, Tfe3 is exclusively cytoplasmic (Figure S4C). We therefore examined Tfe3 localization in the embryo. Tfe3 was localized in the nuclei of both inner cell mass (ICM) (Figure 4D, arrowhead) and trophectoderm (TE) cells at embryonic day (E) 3.5. At E4.5, Tfe3 could be detected in nuclei and cytoplasm of Nanog-positive epiblast cells (Figure 4E, arrowhead) but was exclusively nuclear in GATA4-positive hypoblast (Figure 4E, open arrowhead) and TE cells. In postimplantation E5.5 embryos, Tfe3 was strongly expressed but excluded from the nucleus in Oct4-positive epiblast cells (Figure 4F, arrowhead), whereas it remained nuclear in the extraembryonic endoderm (Figure 4F, open arrowhead). Thus, nuclear exclusion of Tfe3 in the epiblast correlates with progression from the naive preim-plantation state to the postimpreim-plantation egg cylinder.
Tfe3 Acts in ESC Commitment
To test whether Tfe3 is functionally downstream of Flcn-Fnip1/2, we performed simultaneous siRNA transfections. Knockdown of the pluripotency-associated genes Klf4, Rex1, or Stat3 did not restore commitment in Tcf3-, Flcn-, Fnip1/2-, or Tsc2-depleted cells. However, knockdown of Tfe3 completely suppressed the Flcn, Fnip1/2, and Tsc2 phenotypes although only mildly affecting Tcf3-dependent commitment (Figures 5A andS5A). Conversely, transfection of Flcn or Fnip1/2 siRNAs had little impact in Tfe3 shRNA knockdown cells (Figure S5B). Thus, Tfe3 is epistatic to Flcn, Fnip1/2, and Tsc2.
Upon depletion of Tfe3, we saw no effect on self-renewal in 2i but a reduction in ESC clonogenicity 24 hr after inhibitor with-drawal (Figures 5B andS5C). Interestingly, knockdown of the Tfe3 family member Tfeb had little effect on its own. We conclude that Tfe3 is not essential for naive pluripotency but constrains exit from ESC self-renewal.
To examine whether nuclear Tfe3 is sufficient to prevent loss of pluripotency, we generated ESCs expressing Tfe3 fused to ERT2 to allow for induction of nuclear translocation by Tam. Tfe3 and Tfe3-ERT2 transfectants were withdrawn from 2i in
the absence or presence of Tam and, after 3–4 days, replated in 2i with selection for Rex1 expression. Whereas Tfe3-express-ing cells exhibited only mild commitment defects, Tam-induced Tfe3-ERT2-expressing cells showed significantly reduced exit from the ESC state (Figures 5C and 5D). To address the timing of Tfe3 action, we pulsed exposure to Tam. Induction during the first 24 hr after 2i withdrawal was sufficient to impair commit-ment, whereas later Tam treatment gave weaker phenotypes (Figure 5E). Rex1GFPd2 cells expressing Tfe3-ERT2 maintained GFP after 24 hr in N2B27 in the presence of Tam (data not shown). Even after 76 hr, Rex1GFPd2 expression was main-tained in a substantial subset of cells (10%–20%) (data not shown). To test whether this heterogeneous response is due to Tfe3-ERT2 expression variation within the pool of transfec-tants or due to variable phenotypic penetrance, we picked clones with defined expression levels. Indeed, higher expres-sion and, consequently, more nuclear Tfe3 (Figure S5D) corre-lated with a higher frequency of GFP maintenance after 3 days of differentiation (Figure S5E). These data are consistent with nuclear Tfe3 inhibiting initial events in the exit from naive pluripotency.
To test for possible crosstalk between the Flcn-Fnip1/2-Tfe3 axis and the differentiation-promoting activities of GSK3 or MAPK signaling, we analyzed downstream targets. FGF4 induces Etv4 and Spry4 (Lanner et al., 2010), whereas GSK3 inhibition derepresses Axin2 and Cdx1 (Martello et al., 2012;
Wray et al., 2011). Upon removal of the respective inhibitor, mRNA levels changed similarly to those in controls in Flcn shRNA, Flcn knockout, or ectopic nuclear Tfe3 cell lines ( Fig-ure S5F and data not shown). Thus, Tfe3 is unlikely to generally inhibit GSK3 or MAPK signaling. Overlap between pathways regulating ESC self-renewal is well documented (Martello et al., 2012;Nichols and Smith, 2012;Niwa et al., 2009;Wray et al., 2010), and we therefore tested whether Flcn and Tfe3 could compensate for GSK3 or MAPK inhibition. Flcn shRNA or ectopic nuclear Tfe3 ESC lines cultured in the presence of either inhibitor alone maintained higher Rex1GFPd2 expression compared to the respective controls even after serial passaging (Figure S5G). This indicates that Flcn-Fnip1/2-Tfe3 operates independently of GSK3 or MAPK signaling as a third module to enable ESC differentiation.
ESC differentiation products can be converted into EpiSCs by continuous exposure to FGF2 and Activin A (Guo et al., 2009). After three passages in these culture conditions, both Tfe3-ERT2-expressing and Flcn knockdown cell populations acquired a transcriptional signature similar to that of EpiSCs (Figure S5H, upper panel). However, a small percentage of cells remained clonogenic in 2i (Figure S5H, lower panel), suggesting that nuclear Tfe3 reduces transition into a postimplantation cell fate even in a prolonged and highly inductive context.
Ablation of signals mediating exit from the ESC state during differentiation might also facilitate reprogramming. However, neither depletion of Flcn, Fnip1/2, or Tsc2 nor nuclear transloca-tion of Tfe3 enhanced reprogramming of O4GIP EpiSCs into naive pluripotent ESCs, even in sensitized backgrounds overex-pressing Klf4 (Guo et al., 2009) and Nanog (Silva et al., 2009), or hyperactive LIF-STAT3 signaling (Yang et al., 2010) (Figures S5I and S5J and data not shown). Thus, Flcn-Fnip1/2-Tfe3 acts
specifically in exit from pluripotency during differentiation and not in acquisition of pluripotency during reprogramming.
Identification of Tfe3 Downstream Targets
To explore the molecular function of Tfe3 downstream of Flcn, we performed chromatin immunoprecipitation coupled to deep sequencing (ChIP-Seq). We identified 6,789 peaks in control and 8,941 peaks in Flcn shRNA cell lines. Cluster analysis showed a high overlap between both Tfe3 ChIP data sets ( Fig-ures S6A and S6B), and de novo motif recognition identified a modified E box in more than 80% of bound loci in both experi-ments (Figure S6C). This suggests that Flcn-Fnip1/2 regulates the amount of chromatin-bound Tfe3 rather than binding speci-ficity. Consistent with this, normalized read numbers per peak region were greater in Flcn shRNA compared to control cells (Figure S6D).
We compared Tfe3 binding with published ESC ChIP data sets (Martello et al., 2012). Cluster analysis of peak regions ( Fig-ure S6A) and associated genes (Figure S6B) showed that the
Tfe3-binding profile is distinct and well separated from that of a group containing the bHLH transcription factor n-Myc but closer to the core pluripotency factor family including Oct4, Nanog, and Sox2. The majority of sites occupied by Oct4, Sox2, Nanog, and Tcf3 are distant from sites bound by Tfe3 ( Fig-ure S6E). However, a significant fraction of common target genes of Oct4, Sox2, Nanog, and Tcf3 are associated with Tfe3 sites (Figure S6F). This group is more likely transcriptionally up- or downregulated during ESC differentiation compared with genes exclusively bound by Tfe3 (Figure S6G), suggesting a functional relationship between Tfe3 and the pluripotency network.
Because pluripotency transcription factors are strongly relayed onto themselves and exhibit pronounced colocation (Chen et al., 2008b), we examined the overlap of Tfe3, Oct4, Sox2, Nanog, and Tcf3 peaks (Table S1). Co-occupied loci included putative cis-regulatory elements in the pivotal pluripo-tency regulators Klf2, Esrrb (Figure 6A), and Tbx3 (data not shown), suggesting that these may be nodes where Tfe3 con-nects to the pluripotent circuitry.
C D E
Figure 5. Tfe3 Is Required and Sufficient to Impair ESC Commitment
(A) Commitment of O4GIP cells transfected with the indicated siRNA combinations assayed after 72 hr of differentiation by reapplying 2i culture conditions and Oct4-expression selection. Exit from pluripotency was quantified with a cell-viability assay and normalized to negative siRNA treatment. Average and SD are of two technical replicates.
(B) Rex1GFPd2 cells were transfected with indicated siRNAs and exposed for 24 hr to the indicated culture conditions, and ESCs were quantified by replating single cells in 2i with Rex1-expression selection. Average clonogenicity relative to negative siRNA and SD are of four independent experiments.
(C and D) Rex1GFPd2 cells expressing indicated constructs were differentiated in the absence or presence of 0.1mM Tam for 100 hr and replated in 2i with Rex1 selection (five times fewer cells were replated for cells kept in 2i), and resulting ESC colonies were visualized with AP (C). Cells differentiated for 72 hr were replated at clonal density, and resulting colonies arising from uncommitted cells stained for AP (D). Average clonogenicity relative to number of plated cells and SD are of two independent experiments.
(E) O4GIP ESCs expressing an empty vector or Tfe3-ERT2 were differentiated in the indicated intervals with 0.1mM Tam and EtOH and switched back to 2i with puromycin selection, and ESCs were quantified with a cell-viability assay. Average fold changes over empty vector control at 72 hr EtOH and SD are from two technical replicates.
To assess functionality of Tfe3 binding, we analyzed mRNA levels after 3 hr of Tam addition to Tfe3-ERT2-expressing cells. We included ApoE and Trpm1, which are bound by Tfe3 (data not shown) and have been described to be regulated by Flcn (Hong et al., 2010) and the Tfe3 homolog MiTF (Miller et al., 2004), respectively. In 2i, we detected induction of Trpm1 and ApoE and a modest but significant upregulation of Esrrb, though not Tbx3 or Klf2 (Figure 6B). Although only about 2-fold, the
upregulation of Esrrb is notable on top of the induction by GSK3 inhibition (Martello et al., 2012). In fact, less than 2-fold overexpression of Esrrb was sufficient to impede exit from plu-ripotency (Figures S6H and S6I). To test whether Flcn regulates this activity of Tfe3, we analyzed target genes after Flcn siRNA transfection. Indeed, Esrrb, ApoE, and Trpm1 were upregulated after Flcn knockdown but not after codepletion of Tfe3. In contrast, an unrelated Tfe3-binding target, Smad7, was not
Figure 6. Genome-wide Tfe3-Target Determination IdentifiesEsrrb as a Downstream Effector of Flcn-Fnip1/2-Tfe3
(A) Gene tracks of loci identified by Tfe3 ChIP-Seq with control (ctrl.2), Flcn shRNA (Flcn.4), and Tfe3 shRNA (Tfe3.3) cell lines, overlaid with Nanog-, Oct4- and Tcf3-bound regions.
(B) ESCs expressing indicated constructs were treated for 3 hr with 0.1mM Tam in 2i. Average mRNA fold changes relative to EtOH treatment and SD are from two independent experiments with three different cell lines per genotype. (*) indicates Student’s t test values < 0.005.
(C) Average mRNA changes in ESCs transfected with the indicated siRNA combinations. Average relative expression normalized to no and negative siRNA treatments and SD are of three independent experiments. (*) indicates Student’s t test values < 0.05.
(D) O4GIP ESCs expressing Tfe3-ERT2 were transfected with indicated siRNAs, differentiated for 3 days in the presence of 0.1mM Tam or EtOH, and switched back to 2i with puromycin selection, and remaining ESC colonies were quantified with a cell-viability assay. Average fold changes over negative siRNA-trans-fected, EtOH-treated cells and SD are from two independent experiments. (**) and (*) indicate Student’s t test values < 0.001 and 0.03, respectively. (E) Commitment of O4GIP cells transfected with indicated siRNA combinations, including two independent Esrrb siRNAs. Exit from pluripotency was quantified with a cell-viability assay and normalized to negative siRNA treatment. Average and SD are of two technical replicates.
Esrrb siRNAs together with Flcn or Fnip1/2 depletion restored commitment (Figure 6E). Conversely, nuclear Tfe3 failed to inhibit exit from the ES cell state in Esrrb knockout cells ( Fig-ure S6J). These results suggest that Esrrb acts epistatically to Flcn-Fnip1/2-Tfe3. We note, however, that knockdown of Esrrb alone destabilized self-renewal (Figures 6D and 6E), consistent with its pivotal role in the pluripotency circuit (Nichols and Smith, 2012).
Nuclear Tfe3 Maintains an ESC State
Tfe3-ERT2-expressing ESCs differentiated into neural progeni-tors and neurons in N2B27 (Figure 7A, left panel). In the presence of Tam, however, many cells resisted differentiation. These cultures acquired stable morphology within 2–4 passages ( Fig-ure 7A, right panel) and could be propagated in N2B27 contain-ing Tam for at least 25 passages. Tfe3-ERT2-Tam (TET) cells derived from Rex1GFPd2 and O4GIP ESCs maintained GFP expression in approximately 60%–90% and 90% of cells in the population, respectively (Figure S7A and data not shown). Tran-script profiling showed that, with the exception of Nanog, TET cells maintained expression of pluripotency genes at levels similar to those in 2i cells (Figure 7B) and were distinct from EpiSCs (Figure S7B). Flcn knockout but not wild-type or hetero-zygous ESCs could be similarly sustained in N2B27 alone, although expression levels of pluripotency genes appeared lower (Figure S7C). The differentiation markers FGF5, GATA3, and GATA4 were upregulated in both cases (Figures 7B,S7C, and S7D). To test whether this is because of ongoing differenti-ation in the cultures, we sorted GFP-high and -low Rex1GFPd2 TET cells. Indeed, GFP-low cells expressed lower levels of plu-ripotency and higher levels of differentiation markers compared to GFP-high cells (Figure S7E), indicating that TET cell cultures show heterogeneity, similar to ESCs in serum/LIF (Marks et al., 2012).
Withdrawal of Tam caused downregulation of pluripotency gene expression, whereas the neural progenitor markers Sox1, Pax6, Hes5, and Ascl1 were induced (Figures 7C and 7D). Both Flcn knockout cells and TET cells in the presence of Tam self-renewed at single-cell level, generating AP-positive colonies with a skirt of negative cells (mixed) but very few AP-negative, differentiated colonies (Figures 7E andS7F). Addition of a Janus kinase (JAK) inhibitor did not alter colony composition, indicating that LIF-JAK signaling is not required for TET cell self-renewal.
Knockdown of Esrrb, but not Klf2, Klf4, or Tbx3, in Rex1GFPd2 TET cells caused downregulation of GFP expression ( Fig-ure S7G), and replating at clonal density showed impaired self-renewal of Esrrb-depleted cells (Figure 7F). This is further evidence that Esrrb is specifically required for TET cell
self-for high GFP and after two further passages in Tam injected into mouse blastocysts. Resulting embryos showed widespread contribution of GFP-positive cells at E11.5 (Figure 7G). Further-more, two independently generated TET cell lines contributed to adult tissues. Extensive coat color chimerism was obtained with or without pretreatment in 2i in 4/5 or 4/9 and 6/7 or 2/7 mice, respectively (Figures 7H andS7H and data not shown).
DISCUSSION
Exit from the ESC state requires dissolution of the core pluripo-tency transcription-factor circuit, but how the differentiation machinery breaks down this network is unclear. Recovery of several activators in the GSK3 and MAPK pathways (Figure 1D) indicated that our screen was effective in identifying drivers of commitment. GSK3, via b-catenin and Tcf3, represses the orphan nuclear receptor Esrrb, a central member of the pluripo-tent circuitry (Festuccia et al., 2012;Ivanova et al., 2006;Martello et al., 2012;Percharde et al., 2012). Besides Tcf3 and the GSK3-scaffold gene APC, we recovered CtBP2, a Tcf3 interactor that may act as a transcriptional corepressor (Brannon et al., 1999;
Tarleton and Lemischka, 2010). Autocrine FGF4 signaling drives ESC differentiation, and we recovered two kinases, KRas and Raf1, in this pathway as well as the MAPK-scaffold protein Mapksp1/Mp1 (Westerman et al., 2011). Mp1 is required for MAPK signaling at late endosomes (Teis et al., 2002), and we also identified Hgs/Hrs, an ESCRT-0 complex member important for FGF receptor endocytosis and signaling (Chanut-Delalande et al., 2010).
We found that efficient exit from naive pluripotency requires components not previously implicated in ESC biology, namely Flcn-Fnip1/2 and the mTOR regulators Tsc1 and Tsc2, that converge on relocalization of the bHLH transcription factor Tfe3. Conversely, enforced nuclear Tfe3 prevents commitment and maintains an ESC state. Interestingly, Tsc1/2 and Flcn are tumor suppressors, whereas Tfe3 is a protooncogene. Tfe3-gene fusions can be causal for renal cell carcinomas, which also occur in tuberosclerosis due to mutations in Tsc1 or Tsc2 and in Bird-Hogg-Dube´ (BHD) syndrome due to Flcn mutations (Armah and Parwani, 2010;Linehan et al., 2010). There are differ-ences in the histopathological classifications of these tumors, but there may be a common underlying pathway. Consistent with this, nuclear Tfe3 has been described in BHD kidney can-cers (Hong et al., 2010).
The Flcn Pathway
The primary sequence of Flcn does not feature recognizable domains. A Flcn protein complex purified from HEK293 cells
Figure 7. Nuclear Tfe3 Maintains an ESC State
(A) Tfe3-ERT2-expressing cells differentiated into neurons (arrowhead) in the absence (left panel) but retained undifferentiated morphology and could be serially passaged in the presence of 0.1mM Tam (right panel).
(B) mRNA expression in Rex1GFPd2 and O4GIP TET cells relative to ESCs in 2i conditions. Average and SD are of three independent experiments. (C) mRNA expression in Rex1GFPd2 TET cells after Tam withdrawal relative to presence of Tam. Average and SD are of two technical replicates. (D) Immunohistochemistry in Rex1GFPd2 TET cells treated for 6 days with and without 0.1mM Tam.
(E) TET cells were plated at single-cell density in indicated culture conditions. Resulting colonies were stained for AP (lower panel) and quantified. Average numbers relative to Tam and SD are from two independent experiments (left bar: Rex1GFPd2, right bar: O4GIP).
(F) Rex1GFPd2 TET cells were transfected with indicated siRNAs and replated after 2 days at clonal density in the presence of Tam. Average clonogenicity relative to negative siRNA and SD are of two independent experiments.
(G) Rex1GFPd2 TET cells stably transfected with a GFP-expressing plasmid were microinjected into blastocysts, and resulting embryos analyzed at E11.5. Widespread contribution was detected in 4/7 embryos. For comparison, a non-GFP-expressing embryo (arrowhead) of the same litter is shown.
Fnip1-deficient mice are viable (Baba et al., 2012;Park et al., 2012), suggesting redundancy with Fnip2. However, Fnip1 and conditional Flcn mutants show similar defects in B cell develop-ment (Baba et al., 2012), indicating that requirement for Fnip2 is facultative and cell type dependent.
Depletion of Tsc1 or Tsc2, similar to Flcn-Fnip1/2, induced nuclear accumulation of Tfe3 and impaired exit from pluripo-tency. Epistasis experiments with rapamycin and Tfe3 are consistent with mTOR acting upstream of and/or in parallel with Flcn-Fnip1/2 to inhibit Tfe3-dependent ESC commitment. Phosphorylation of Flcn downstream of mTOR has been described (Baba et al., 2006;Piao et al., 2009), suggesting a linear relationship between Tsc1/2, mTOR, and Flcn-Fnip1/2 (Figure 7I). However, the identity of the kinase and how phos-phorylation affects Flcn remains to be determined. In vivo, Tsc1 and Tsc2 knockout mice show defects after gastrulation at around E10.0 (Kobayashi et al., 1999,2001), whereas Flcn mutants are affected by E6.0 (Hasumi et al., 2009). Tsc mutant embryos feature phenotypic variability and developmental delay, though, which may suggest earlier defects that are compensated for by regulative mechanisms.
We postulate that Tsc1/2 and Flcn-Fnip1/2 phenotypes in ESCs are mediated by nuclear translocation and activation of Tfe3 (Figure 7I). Depletion of Tfe3 accelerated ESC commitment, indicating that Tfe3 resists breakdown of the pluripotency network. However, Tfe3 mutant mice are viable and fertile (Steingrimsson et al., 2002). MiTF, Tfe3, and Tfeb can interact with each other (Hemesath et al., 1994), and functional redun-dancy has been reported (Huan et al., 2006; Steingrimsson et al., 2002). In fact, although Tfeb was not essential for Tsc2 or Flcn-Fnip1/2 phenotypes (Figures S5A and6E), its overex-pression impaired differentiation (data not shown). Thus, redundancy could explain why Tfe3 mutants have no overt developmental phenotype. It should also be noted that several factors important for in vitro ESC self-renewal are dispensable in the embryo (Nichols and Smith, 2012).
Regulation of Pluripotency
However, early lethality of Flcn mutant embryos is consistent with a role of the Flcn-Fnip1/2-Tfe3 module in progression of pluripotency in vivo. We observed a change in subcellular local-ization of Tfe3, suggesting that Flcn-Fnip1/2 is inactive in the
Tfe3 similarly translocates into the cytoplasm at the onset of ESC differentiation and is cytoplasmic in EpiSCs. These obser-vations suggest that the Flcn-Fnip1/2-Tfe3 module acts as a cell-fate switch that steers naive early epiblast and ESCs into a primed postimplantation state in vitro and in vivo.
Flcn-Fnip1/2-Tfe3 activity requires an ESC-specific context, as nuclear translocation of Tfe3 or knockdown of Tsc2 and Flcn-Fnip1/2 in EpiSCs did not induce naive pluripotency. Similarly, ectopic activation of Tfe3 efficiently impeded ESC dif-ferentiation in the first 24 hr after inhibitor withdrawal, but less so afterward (Figure 5E).
The ability of nuclear Tfe3 to maintain an ESC state in the complete absence of pro-self-renewal or antidifferentiation sig-nals is unprecedented and implies potent input to the transcrip-tional circuitry sustaining self-renewal (Nichols and Smith, 2012). Functional analysis indicates that transcriptional upregu-lation of Esrrb is a relevant output of nuclear Tfe3. It is likely that there are additional Tfe3 targets because TET cells self-renew without requirement for either MAPK or GSK3 inhibition, whereas forced expression of Esrrb confers independence only from the latter (Martello et al., 2012). Nonetheless, it is striking that transcription of Esrrb is regulated by Flcn-Fnip1/ 2-Tfe3, GSK3/b-catenin/Tcf3 (Martello et al., 2012), and Nanog (Festuccia et al., 2012) (Figure 7I). We surmise that repression by Tcf3 combined with withdrawal of Tfe3 and downregulation of Nanog act to ensure prompt and efficient silencing of this pivotal pluripotency factor at the onset of differentiation. It remains to be tested whether distinct regulatory inputs act in parallel or in sequence, but it is striking that absence of Tcf3 does not infringe development prior to gastrulation (Merrill et al., 2004), suggesting compensatory dependence of Esrrb on Nanog or Tfe3 in vivo.
In summary, this study has identified a module that was previ-ously undescribed in ESCs and mediates a switch from naive to primed pluripotent states. Our findings suggest that nuclear exclusion of Tfe3 acts as a rheostat for controlled disequilibra-tion of the regulatory network that establishes ground-state plu-ripotency in the embryo and sustains ESCs in vitro.
EXPERIMENTAL PROCEDURES
Full details are provided in theExtended Experimental Procedures.
(H) Passage 8 Rex1GFPd2 TET cells were cultured for 2 days in 2i and injected into C57BL/6 blastocysts. Contribution of the TET cell agouti gene to coat color is visible against black host fur.
(I) Schematic representation of the Flcn-Fnip1/2 pathway and combinatorial inputs into Esrrb transcription. (..) denotes additional Tfe3 targets that contribute to self-renewal.
Mice and Embryos
Chimeras were generated by microinjection into C57BL/6 blastocysts. Animal experiments were approved by the University of Cambridge Ethical Review Committee and authorized by the UK Home Office.
Cell Culture
ESCs were maintained in serum-free N2B27 medium containing 2i and, where indicated, LIF (Wray et al., 2011;Ying et al., 2008). EpiSCs were cultured as described (Guo et al., 2009).
Commitment Assay
ESCs were plated on gelatin-coated plates at 1.53 104 cells/cm2
. The next day, cells were washed with PBS (PAA) and medium changed to N2B27 without 2i or LIF. Commitment of O4GIP ESCs was determined by restoring 2i plus 1mg/ml puromycin (Sigma). Uncommitted ESCs were quantified after 2–3 days by adding Alamar Blue (Invitrogen), diluted 1:10 in 2i medium, and read out on a SpectraMax M2 (Molecular Devices). Commitment of other genotypes was quantified by replating on laminin-coated plates in ESC medium. For Rex1GFPd2 cells, 10mg/ml blasticidin (PAA) was added to select for Rex1 expression.
siRNA Transfection
Cells were reverse transfected with 16.7 nM siRNA in 2i or FGF2/Activin A containing N2B27 medium with transfection mixes prepared with RNAiMax or Lipofectamine 2000 (Invitrogen) in OptiMEM (Invitrogen). All siRNAs with the exception of Tcf7l1/Tcf3 (Dharmacon, L-048614-01-0005) and Tbx3 (Ambion, 223884 and 223885) were purchased from QIAGEN (FlexiTube GeneSolution siRNA pools; AllStars Negative Control siRNA, 1027281; GFP-22 siRNA, 1022064). Knockdown of Tcf3 is included as a reference.
ACCESSION NUMBERS
ChIP-Seq data sets have been deposited in the GEO database with the ID number GSE39815.
SUPPLEMENTAL INFORMATION
Supplemental Information includes Extended Experimental Procedures, seven figures, and two tables and can be found with this article online athttp://dx.doi. org/10.1016/j.cell.2013.03.012.
ACKNOWLEDGMENTS
We thank Tu¨zer Kalkan and Graziano Martello for help, advice, and reagents; Raja Kittappa, Martin Leeb, and Harry Leitch for comments; Ravi Nookala for discussion; Ge Guo for providing OEC-2 EpiSCs; and Sandra Go´mez-Lo´pez and Jose Silva for sharing CreERT2, pDONR221-Klf4, and -Nanog plasmids. We are grateful to Laura Schmidt for generously providing Flcn antibodies and mice; Kathryn Lilley, Mike Deery, and Svenja Hesters of the Cambridge Centre for Proteomics for performing and analyzing mass spectrometry; Rachael Walker for flow cytometry; William Mansfield and Charles-Etienne Dumeau for blastocyst injections; Samuel Jameson and team for husbandry; Peter Humphreys for image quantification; Maike Paramor for ChIP-Seq library generation; and Rosalind Drummond and Melanie Rittirsch for technical assis-tance. This study was funded by the Medical Research Council, Wellcome Trust, and the European Commission FP7 project SyBOSS (242129). J.B. is recipient of a Human Frontier Science Program fellowship (LT00686/2007-L), P.J.P. is a Pew Scholar in Biomedical Sciences, and A.S. is a Medical Research Council Professor. J.B. performed and interpreted experiments, J.N. performed embryo immunohistochemistry and ESC derivations, S.D. per-formed bioinformatical analysis, and P.D.C. and P.J.P. provided and assisted with the siRNA-screening platform. A.S. supervised the study and wrote the paper with J.B. Received: August 3, 2012 Revised: January 14, 2013 Accepted: March 7, 2013 Published: April 11, 2013 REFERENCES
Armah, H.B., and Parwani, A.V. (2010). Xp11.2 translocation renal cell carci-noma. Arch. Pathol. Lab. Med. 134, 124–129.
Baba, M., Hong, S.B., Sharma, N., Warren, M.B., Nickerson, M.L., Iwamatsu, A., Esposito, D., Gillette, W.K., Hopkins, R.F., 3rd, Hartley, J.L., et al. (2006). Folliculin encoded by the BHD gene interacts with a binding protein, FNIP1, and AMPK, and is involved in AMPK and mTOR signaling. Proc. Natl. Acad. Sci. USA 103, 15552–15557.
Baba, M., Keller, J.R., Sun, H.W., Resch, W., Kuchen, S., Suh, H.C., Hasumi, H., Hasumi, Y., Kieffer-Kwon, K.R., Gonzalez, C.G., et al. (2012). The folliculin-FNIP1 pathway deleted in human Birt-Hogg-Dube´ syndrome is required for murine B-cell development. Blood 120, 1254–1261.
Brannon, M., Brown, J.D., Bates, R., Kimelman, D., and Moon, R.T. (1999). XCtBP is a XTcf-3 co-repressor with roles throughout Xenopus development. Development 126, 3159–3170.
Brons, I.G., Smithers, L.E., Trotter, M.W., Rugg-Gunn, P., Sun, B., Chuva de Sousa Lopes, S.M., Howlett, S.K., Clarkson, A., Ahrlund-Richter, L., Pedersen, R.A., and Vallier, L. (2007). Derivation of pluripotent epiblast stem cells from mammalian embryos. Nature 448, 191–195.
Chanut-Delalande, H., Jung, A.C., Baer, M.M., Lin, L., Payre, F., and Affolter, M. (2010). The Hrs/Stam complex acts as a positive and negative regulator of RTK signaling during Drosophila development. PLoS ONE 5, e10245. Chen, J., Futami, K., Petillo, D., Peng, J., Wang, P., Knol, J., Li, Y., Khoo, S.K., Huang, D., Qian, C.N., et al. (2008a). Deficiency of FLCN in mouse kidney led to development of polycystic kidneys and renal neoplasia. PLoS ONE 3, e3581. Chen, X., Xu, H., Yuan, P., Fang, F., Huss, M., Vega, V.B., Wong, E., Orlov, Y.L., Zhang, W., Jiang, J., et al. (2008b). Integration of external signaling pathways with the core transcriptional network in embryonic stem cells. Cell
133, 1106–1117.
Doble, B.W., Patel, S., Wood, G.A., Kockeritz, L.K., and Woodgett, J.R. (2007). Functional redundancy of GSK-3alpha and GSK-3beta in Wnt/beta-catenin signaling shown by using an allelic series of embryonic stem cell lines. Dev. Cell 12, 957–971.
Festuccia, N., Osorno, R., Halbritter, F., Karwacki-Neisius, V., Navarro, P., Colby, D., Wong, F., Yates, A., Tomlinson, S.R., and Chambers, I. (2012). Esrrb is a direct Nanog target gene that can substitute for Nanog function in plurip-otent cells. Cell Stem Cell 11, 477–490.
Guo, G., Yang, J., Nichols, J., Hall, J.S., Eyres, I., Mansfield, W., and Smith, A. (2009). Klf4 reverts developmentally programmed restriction of ground state pluripotency. Development 136, 1063–1069.
Hartman, T.R., Nicolas, E., Klein-Szanto, A., Al-Saleem, T., Cash, T.P., Simon, M.C., and Henske, E.P. (2009). The role of the Birt-Hogg-Dube´ protein in mTOR activation and renal tumorigenesis. Oncogene 28, 1594–1604. Hasumi, H., Baba, M., Hong, S.B., Hasumi, Y., Huang, Y., Yao, M., Valera, V.A., Linehan, W.M., and Schmidt, L.S. (2008). Identification and characterization of a novel folliculin-interacting protein FNIP2. Gene 415, 60–67.
Hasumi, Y., Baba, M., Ajima, R., Hasumi, H., Valera, V.A., Klein, M.E., Haines, D.C., Merino, M.J., Hong, S.B., Yamaguchi, T.P., et al. (2009). Homozygous loss of BHD causes early embryonic lethality and kidney tumor development with activation of mTORC1 and mTORC2. Proc. Natl. Acad. Sci. USA 106, 18722–18727.
Hayashi, K., Ohta, H., Kurimoto, K., Aramaki, S., and Saitou, M. (2011). Recon-stitution of the mouse germ cell specification pathway in culture by pluripotent stem cells. Cell 146, 519–532.
Hemesath, T.J., Steingrı´msson, E., McGill, G., Hansen, M.J., Vaught, J., Hodg-kinson, C.A., Arnheiter, H., Copeland, N.G., Jenkins, N.A., and Fisher, D.E.
Jaenisch, R., and Young, R. (2008). Stem cells, the molecular circuitry of pluripotency and nuclear reprogramming. Cell 132, 567–582.
Kaji, K., Caballero, I.M., MacLeod, R., Nichols, J., Wilson, V.A., and Hendrich, B. (2006). The NuRD component Mbd3 is required for pluripotency of embry-onic stem cells. Nat. Cell Biol. 8, 285–292.
Kobayashi, T., Minowa, O., Kuno, J., Mitani, H., Hino, O., and Noda, T. (1999). Renal carcinogenesis, hepatic hemangiomatosis, and embryonic lethality caused by a germ-line Tsc2 mutation in mice. Cancer Res. 59, 1206–1211. Kobayashi, T., Minowa, O., Sugitani, Y., Takai, S., Mitani, H., Kobayashi, E., Noda, T., and Hino, O. (2001). A germ-line Tsc1 mutation causes tumor devel-opment and embryonic lethality that are similar, but not identical to, those caused by Tsc2 mutation in mice. Proc. Natl. Acad. Sci. USA 98, 8762–8767. Kunath, T., Saba-El-Leil, M.K., Almousailleakh, M., Wray, J., Meloche, S., and Smith, A. (2007). FGF stimulation of the Erk1/2 signalling cascade triggers tran-sition of pluripotent embryonic stem cells from self-renewal to lineage commit-ment. Development 134, 2895–2902.
Lanner, F., Lee, K.L., Sohl, M., Holmborn, K., Yang, H., Wilbertz, J., Poellinger, L., Rossant, J., and Farnebo, F. (2010). Heparan sulfation-dependent fibro-blast growth factor signaling maintains embryonic stem cells primed for differ-entiation in a heterogeneous state. Stem Cells 28, 191–200.
Laplante, M., and Sabatini, D.M. (2012). mTOR signaling in growth control and disease. Cell 149, 274–293.
Linehan, W.M., Srinivasan, R., and Schmidt, L.S. (2010). The genetic basis of kidney cancer: a metabolic disease. Nat Rev Urol 7, 277–285.
Marks, H., Kalkan, T., Menafra, R., Denissov, S., Jones, K., Hofemeister, H., Nichols, J., Kranz, A., Stewart, A.F., Smith, A., and Stunnenberg, H.G. (2012). The transcriptional and epigenomic foundations of ground state plurip-otency. Cell 149, 590–604.
Martello, G., Sugimoto, T., Diamanti, E., Joshi, A., Hannah, R., Ohtsuka, S., Go¨tt-gens, B., Niwa, H., and Smith, A. (2012). Esrrb is a pivotal target of the Gsk3/Tcf3 axis regulating embryonic stem cell self-renewal. Cell Stem Cell 11, 491–504. Merrill, B.J., Pasolli, H.A., Polak, L., Rendl, M., Garcı´a-Garcı´a, M.J., Anderson, K.V., and Fuchs, E. (2004). Tcf3: a transcriptional regulator of axis induction in the early embryo. Development 131, 263–274.
Miller, A.J., Du, J., Rowan, S., Hershey, C.L., Widlund, H.R., and Fisher, D.E. (2004). Transcriptional regulation of the melanoma prognostic marker melasta-tin (TRPM1) by MITF in melanocytes and melanoma. Cancer Res. 64, 509–516. Nichols, J., and Smith, A. (2009). Naive and primed pluripotent states. Cell Stem Cell 4, 487–492.
Nichols, J., and Smith, A. (2011). The origin and identity of embryonic stem cells. Development 138, 3–8.
Nichols, J., and Smith, A. (2012). Pluripotency in the embryo and in culture. Cold Spring Harb. Perspect. Biol. 4, a008128.
Nickerson, M.L., Warren, M.B., Toro, J.R., Matrosova, V., Glenn, G., Turner, M.L., Duray, P., Merino, M., Choyke, P., Pavlovich, C.P., et al. (2002). Muta-tions in a novel gene lead to kidney tumors, lung wall defects, and benign tumors of the hair follicle in patients with the Birt-Hogg-Dube´ syndrome. Can-cer Cell 2, 157–164.
Niwa, H., Ogawa, K., Shimosato, D., and Adachi, K. (2009). A parallel circuit of LIF signalling pathways maintains pluripotency of mouse ES cells. Nature 460, 118–122.
Piao, X., Kobayashi, T., Wang, L., Shiono, M., Takagi, Y., Sun, G., Abe, M., Hagiwara, Y., Zhang, D., Okimoto, K., et al. (2009). Regulation of folliculin (the BHD gene product) phosphorylation by Tsc2-mTOR pathway. Biochem. Biophys. Res. Commun. 389, 16–21.
Silva, J., Nichols, J., Theunissen, T.W., Guo, G., van Oosten, A.L., Barrandon, O., Wray, J., Yamanaka, S., Chambers, I., and Smith, A. (2009). Nanog is the gateway to the pluripotent ground state. Cell 138, 722–737.
Steingrimsson, E., Tessarollo, L., Pathak, B., Hou, L., Arnheiter, H., Copeland, N.G., and Jenkins, N.A. (2002). Mitf and Tfe3, two members of the Mitf-Tfe family of bHLH-Zip transcription factors, have important but functionally redundant roles in osteoclast development. Proc. Natl. Acad. Sci. USA 99, 4477–4482.
Takagi, Y., Kobayashi, T., Shiono, M., Wang, L., Piao, X., Sun, G., Zhang, D., Abe, M., Hagiwara, Y., Takahashi, K., and Hino, O. (2008). Interaction of folliculin (Birt-Hogg-Dube´ gene product) with a novel Fnip1-like (FnipL/Fnip2) protein. Oncogene 27, 5339–5347.
Tang, F., Barbacioru, C., Bao, S., Lee, C., Nordman, E., Wang, X., Lao, K., and Surani, M.A. (2010). Tracing the derivation of embryonic stem cells from the inner cell mass by single-cell RNA-Seq analysis. Cell Stem Cell 6, 468–478. Tarleton, H.P., and Lemischka, I.R. (2010). Delayed differentiation in embry-onic stem cells and mesodermal progenitors in the absence of CtBP2. Mech. Dev. 127, 107–119.
Teis, D., Wunderlich, W., and Huber, L.A. (2002). Localization of the MP1-MAPK scaffold complex to endosomes is mediated by p14 and required for signal transduction. Dev. Cell 3, 803–814.
Tesar, P.J., Chenoweth, J.G., Brook, F.A., Davies, T.J., Evans, E.P., Mack, D.L., Gardner, R.L., and McKay, R.D. (2007). New cell lines from mouse epiblast share defining features with human embryonic stem cells. Nature
448, 196–199.
Westerman, B.A., Braat, A.K., Taub, N., Potman, M., Vissers, J.H., Blom, M., Verhoeven, E., Stoop, H., Gillis, A., Velds, A., et al. (2011). A genome-wide RNAi screen in mouse embryonic stem cells identifies Mp1 as a key mediator of differentiation. J. Exp. Med. 208, 2675–2689.
Wray, J., Kalkan, T., and Smith, A.G. (2010). The ground state of pluripotency. Biochem. Soc. Trans. 38, 1027–1032.
Wray, J., Kalkan, T., Gomez-Lopez, S., Eckardt, D., Cook, A., Kemler, R., and Smith, A. (2011). Inhibition of glycogen synthase kinase-3 alleviates Tcf3 repression of the pluripotency network and increases embryonic stem cell resistance to differentiation. Nat. Cell Biol. 13, 838–845.
Yang, J., van Oosten, A.L., Theunissen, T.W., Guo, G., Silva, J.C., and Smith, A. (2010). Stat3 activation is limiting for reprogramming to ground state plurip-otency. Cell Stem Cell 7, 319–328.
Yang, S.H., Kalkan, T., Morrisroe, C., Smith, A., and Sharrocks, A.D. (2012). A genome-wide RNAi screen reveals MAP kinase phosphatases as key ERK pathway regulators during embryonic stem cell differentiation. PLoS Genet.
8, e1003112.
Ying, Q.L., Stavridis, M., Griffiths, D., Li, M., and Smith, A. (2003). Conversion of embryonic stem cells into neuroectodermal precursors in adherent mono-culture. Nat. Biotechnol. 21, 183–186.
Ying, Q.L., Wray, J., Nichols, J., Batlle-Morera, L., Doble, B., Woodgett, J., Cohen, P., and Smith, A. (2008). The ground state of embryonic stem cell self-renewal. Nature 453, 519–523.
